key: cord-1016012-pggf4wyo
authors: Schmidt, Rose A.; Genois, Rosalie; Jin, Jonathan; Vigo, Daniel; Rehm, Jürgen; Rush, Brian
title: The early impact of COVID-19 on the incidence, prevalence, and severity of alcohol use and other drugs: A systematic review
date: 2021-09-22
journal: Drug Alcohol Depend
DOI: 10.1016/j.drugalcdep.2021.109065
sha: ed190b355b57340898d661801b9294792593808c
doc_id: 1016012
cord_uid: pggf4wyo

BACKGROUND: The aim of this paper was to examine the early impact of COVID-19 on substance use to assess implications for planning substance use treatment and support systems. METHOD: A systematic review of literature published up to March 2021 was conducted to summarize changes in prevalence, incidence, and severity of substance use associated with COVID-19 and the accompanying public health measures, including lockdown, stay-at-home orders, and social distancing. RESULTS: We identified 53 papers describing changes to substance use at the population level. The majority of papers described changes related to alcohol use and most relied on self-reported measures of consumption during the COVID-19 pandemic, compared with pre-pandemic use. There was less evidence to support changes in non-alcohol substance use. In general, risky pre-pandemic alcohol use, caregiving responsibilities, stress, depression, anxiety, and current treatment for a mental disorder were found to be associated with increased substance use. CONCLUSION: This review provides preliminary data on changes in substance use, indicating that certain segments of the population increased their alcohol use early on in the COVID-19 pandemic and may be at greater risk of harm and in need of additional services. There is a need for additional population-level information on substance use to inform evidence-based rapid responses from a treatment system perspective.

In response to the COVID-19 pandemic, governments all over the world implemented physical distancing measures, cancelled large gatherings, and closed businesses and schools as key strategies to manage the spread of infection. As a result, concerns were raised about the mental health impacts of fear, worry, and isolation related to and public health measures (Rajkumar, 2020; Vigo et al., 2020) . Early in the outbreak, high rates of moderate or severe psychological impact as a result of COVID-19 were reported in China (Wang et al., 2020) . Within the general context of mental health concerns, it is critical to bring attention to substance use (SU) and substance use disorders (SUD) given their associated global disease burden (Degenhardt et al., 2018) as well as the unique challenges that a pandemic may present for those using substances and the services they require (Dunlop et al., 2020; Ornell et al., 2020) .

The impacts of pandemics and associated public health measures can increase negative emotions, which have been linked to SU/SUD and return to SUD, even among long-term abstainers (Da et al., 2020; Ornell et al., 2020; Rajkumar, 2020; Sinha et al., 2009) . Early in the pandemic, national and international surveys reported increased rates of SU, particularly alcohol use. Among the respondents to the Global Drug Survey, 56% of respondents increased the number of days on which they drank alcohol since the start of COVID-19 (Winstock, 2020) . Boredom, depression, worry, loneliness, and pandemicrelated anxiety were reported as reasons for increased use. Surveys with national samples reported that 22% of Americans and 19% of Canadians were drinking more alcohol during COVID-19, and 6% of Canadians were smoking more cannabis (CCSA and NANOS Research, 2020; Sharecare, 2020) . However, from the same study, 12% of J o u r n a l P r e -p r o o f Canadians decreased their alcohol use and 4% of Canadians decreased their cannabis use (CCSA and NANOS Research, 2020) . In some high-income countries, especially at the beginning of COVID-19, alcohol sales increased. In the week of March 21, 2020 alcohol sales increased by 67% in the UK and 50% in the U.S., with a greater increase reported for economy brands and larger volume containers (Da et al., 2020; Eley, 2020) . In some jurisdictions, governments have loosened alcohol restrictions and there may be longerterm effects of shifts in alcohol policy that may never be rolled back, including increases in off-site sales and home delivery (Finlay and Gilmore, 2020; Rehm et al., 2020) .

However, in other jurisdictions the opposite trend has occurred, with the implementation of policies to limit the availability of alcohol (Finlay and Gilmore, 2020; Rehm et al., to heavily, but do not report significant problems related to their SU; Tier 3 to 5 include people with increasing intensity of problems related to their SU, comorbidity including mental health diagnoses, and/ or unmet need for services (Rush et al., 2019b) . See Figure   1 for a description of the five tier categories.

[Insert Figure 1] For NBP, it is of considerable interest to measure the impact of COVID-19 and model the population-level impact to adjust resource requirements and service distribution across the full spectrum of community needs. In the context of our current work on treatment systems and NBP, we undertook a systematic review to examine the early impact of the COVID-19 pandemic and associated public health measures on SU and SUD in order to understand how these changes may impact the need for treatment and support systems. We were particularly interested in exploring the impact at different levels of risk and severity in the population.

We conducted a systematic review of electronic databases in May and October of 2020 and March of 2021. We searched Medline, EMBASE, PsychInfo, and CINAHL using relevant Medical Subject Headings (MeSH) and key words corresponding to our PICO tool (See Table 1 ). The complete search strategy can be found in the Appendix. The 4,030 identified papers were first screened by title and abstract using the software Covidence (https://www.covidence.org). Three reviewers (RS, JJ, RG) each coded the same random sample of 100 papers. As an acceptable level of agreement was obtained for this sample (>80%), the subsequent papers were screened by only one reviewer. The full text of the remaining papers was then retrieved and screened for inclusion by two J o u r n a l P r e -p r o o f reviewers (RG, JJ), and any disagreements were resolved by discussion with RS.

Reference lists of all included studies were screened for relevant articles. Two reviewers (RS, JJ) assessed the included studies using the National Institutes of Health quality assessment tools (https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tool) to identify potential biases in the studies related to our research question. Discrepancies on the quality assessment were resolved through discussions with a third reviewer (RG).

The review methods were registered with the International Prospective Register of Systematic Reviews (CRD42020198670, https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=198670). Figure 2 depicts our Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) diagrams of study selection.

[Insert Figure 2 and Table 1] We included peer-reviewed academic journal articles published up to March 2021 in English, French, or Spanish. We included studies that focused on the impact of living in countries affected by COVID-19, including those studying the impact of lockdown, physical distancing, stay-at-home orders, and other public health measures. Included studies had to report changes to SU service needs at a population level, as indicated by changes in prevalence, incidence, severity, or help-seeking behavior. We excluded papers and results related to tobacco use or that focused only on specific sub-populations (e.g., college students, health care workers). While we initially intended to include papers on the impact of previous respiratory pandemics in this review, we identified only a few papers (e.g., (Lau et al., 2005) ) that fit this criterion. As such, we amended our inclusion J o u r n a l P r e -p r o o f criteria to include only papers on the impact of the COVID-19 pandemic. We did not make any exclusion by study design but excluded brief reports without original data.

We identified 53 papers describing the impact of COVID-19 on substance use prevalence at the population level. Table 2 summarizes the details of the included studies and Table 3 summarizes the results and quality assessment. Most papers used a crosssectional design with data collection occurring between March and May of 2020 (n = 49), and therefore the evidence presented describes changes that occurred early in the pandemic, during the first wave of COVID-19 infections.

[Insert Table 2 and Table 3 ]

Thirty-two studies reported percent population changes in alcohol use behavior during the COVID-19 pandemic compared with pre-pandemic behavior. Twenty of these papers included a single question such as, "Since the start of the pandemic, has your alcohol use increased, decreased, or stayed the same?" In general, most papers (n = 24) reported both increased and decreased alcohol use among different segments of the population. An equal number of papers (n =11) reported that more people increased their drinking (Bartoszek et al., 2020; Every-Palmer et al., 2020; Górnicka et al., 2020; Grossman et al., 2020; Huckle et al., 2020; Rolland et al., 2020; Stanton et al., 2020; Tran et al., 2020; Vanderbruggen et al., 2020; Weerakoon et al., 2020; Zajacova et al., 2020) , and decreased their drinking during the pandemic at the population level (Alpers et al., 2021; Chodkiewicz et al., 2020; Constant et al., 2020; Dozois, 2021; Koopmann et al., J o u r n a l P r e -p r o o f 2020; Kriaucioniene et al., 2020; Oksanen et al., 2021; Rodriguez et al., 2020; Scarmozzino and Visioli, 2020; Sidor and Rzymski, 2020; Szajnoga et al., 2020) .

Twenty studies reported a change in frequency, amount, or severity of alcohol use during the COVID-19 pandemic compared to pre-pandemic use (Alpers et al., 2021; Ammar et al., 2020; Barbosa et al., 2020; Błaszczyk-Bębenek et al., 2020; Boschuetz et al., 2020; Callinan et al., 2021; Callinan et al., 2020; Đogaš et al., 2020; Huckle et al., 2020; Jackson et al., 2021; Newby et al., 2020; Niedzwiedz et al., 2021; Oksanen et al., 2021; Rolland et al., 2020; Sallie et al., 2020; Vanderbruggen et al., 2020; Villanueva et al., 2021; Wardell et al., 2020; Weerakoon et al., 2020; Winkler et al., 2020) , two studies compared alcohol use at two or more time points during the pandemic (Killgore et al., 2021; Pollard et al., 2020) , and one study compared the severity of alcohol use in Hubei province, the epicenter of COVID-19, compared with respondents from other provinces in China (Ahmed et al., 2020) .

Four papers reported an increase in number of drinks consumed per day or per occasion during the pandemic (Barbosa et al., 2020; Boschuetz et al., 2020; Vanderbruggen et al., 2020; Winkler et al., 2020) . A multi-country study reported that, overall, the units of alcohol consumed during quarantine decreased across their sample, except among participants from the UK who reported an increase (Sallie et al., 2020) .

Seven studies reported an increase in how often people consumed alcohol during the pandemic (Błaszczyk-Bębenek et al., 2020; Boschuetz et al., 2020; Pollard et al., 2020) , four of which found that despite the increase in frequency there was either no change or a J o u r n a l P r e -p r o o f reduction in drinks consumed on each drinking occasion (Callinan et al., 2021; Callinan et al., 2020; Niedzwiedz et al., 2021; Wardell et al., 2020) . Boschuetz et al. (2020) also reported an increase in frequency of drinking before 5 PM.

Three studies reported an increase in both the proportion of people who did not drink any alcohol and those who used alcohol heavily or frequently (Błaszczyk-Bębenek et al., 2020; Boschuetz et al., 2020; Đogaš et al., 2020) . In these studies, the results indicated that some occasional drinkers shifted into higher or lower risk categories during the pandemic.

Severity: Five studies found that COVID-19 significantly increased AUDIT scores (Ahmed et al., 2020; Jackson et al., 2021; Killgore et al., 2021; Newby et al., 2020) , although in one study this was only significant among women (Boschuetz et al., 2020) . Killgore et al. (2021) found a positive interaction between the length of lockdown and mean AUDIT score. Four studies did not find increased AUDIT scores during the pandemic (Callinan et al., 2021; Oksanen et al., 2021; Sallie et al., 2020; Villanueva et al., 2021) .

Although Oksanen et al. (2021) did not find an overall change in AUDIT scores, among the 37% of the population that did increase their alcohol use, 53.4% of them had hazardous (≥5) AUDIT scores, indicating that those who already drank at risky levels were more likely to increase their use. Similarly, Sallie et al. (2020) found that overall AUDIT scores decreased, but those that did increase their weekly units consumed during COVID-19 had the highest current drinking severity scores, and Weerakoon et al. (2020) found that binge drinkers were more likely to increase alcohol consumption during the pandemic compared with non-binge drinkers (60% vs. 28%).

Binge/ Hazardous Use: Compared to before the pandemic, two studies reported an increased proportion of people reporting binge drinking (Niedzwiedz et al., 2021; Winkler et al., 2020) . However, two studies reported a decrease in the percentage of the population that reported binge-drinking (Ammar et al., 2020; Boschuetz et al., 2020), and Alpers et al. (2021) reported that rates of binge and hazardous drinking were similar to pre-pandemic rates from other studies.

Four studies reported changes in cannabis use behaviors. Rolland et al. (2020) found that among French adults who used cannabis, 60.5% had made changes to their use; 24.3% had moderately increased, 6.9% had increased in a way that was "difficult to control," 22.5% had reduced or stopped without cravings or withdrawal, and 6.7% had reduced or stopped but did experience cravings or withdrawal. Vanderbruggen et al.

(2020) reported that 2.1% of their Belgian sample reported using more cannabis, while 1.1% reported using less, and there was no significant change in number of joints smoked per day. In Canadian samples, Dozois (2021) found that 29% had increased and 48% reported no change in cannabis use, and Zajacova et al. (2020) found that 5.5% reported increased cannabis use, but they did not report the proportion that decreased or stayed the same. Chodkiewicz et al. (2020) found that 1.4% of their Polish sample increased their use of "recreational drugs" and 1.1% decreased their use. A significant majority of their sample, 93.7%, already did not use these drugs. Taylor et al. (2021) found that 31% of their sample who reported drug use prior to the pandemic (16%) had increased their use. J o u r n a l P r e -p r o o f Czeisler et al. (2020) investigated if people in the U.S. had started or increased their SU as a way of coping with COVID-19 stress, including the use of alcohol, legal or illegal drugs, or prescription drugs taken in a way not recommended by a doctor (Czeisler et al., 2020) . They found that 13.3% had started or increased SU to cope with stress, however they did not report the number of people who had reduced their SU (Czeisler et al., 2020) . Similarly, McKnight-Eily et al. (2021) found that 18.2% increased or initiated substance use during the pandemic.

Gender: Thirty studies investigated the impact of gender. Gender differences were noted in nineteen studies (Ahmed et al., 2020; Alpers et al., 2021; Barbosa et al., 2020; Boschuetz et al., 2020; Callinan et al., 2021; Callinan et al., 2020; Đogaš et al., 2020; Ferrante et al., 2020; Gadermann et al., 2021; Garnett et al., 2021; Grossman et al., 2020; Huckle et al., 2020; Killgore et al., 2021; Neill et al., 2020; Niedzwiedz et al., 2021; Pollard et al., 2020; Rodriguez et al., 2020; Rossinot et al., 2020; Sallie et al., 2020) , with no significant differences reported in eleven (Bartoszek et al., 2020; Chodkiewicz et al., 2020; Czeisler et al., 2020; Oksanen et al., 2021; Rolland et al., 2020; Sidor and Rzymski, 2020; Szajnoga et al., 2020; Vanderbruggen et al., 2020; Villanueva et al., 2021; Wardell et al., 2020; Zajacova et al., 2020) . The impact of gender was mixed. Two papers found men increased their drinking more than women (Ferrante et al., 2020; Gadermann et al., 2021) , and that change in drinking amount was positively correlated with being male (Sallie et al., 2020) . However, other studies found that being a man was protective for increased alcohol consumption during the pandemic (OR = 0.71 [95% CI: 0.56. 0.90]; Rossinot et al., 2020) , and that a larger percentage of men had decreased their J o u r n a l P r e -p r o o f alcohol consumption (29.2%) compared with increased consumption (13.1%) or staying the same (12.5%, X 2 p = 0.002; Grossman et al., 2020).

Other studies found significant changes among women's drinking, with more women reporting increased drinking compared with men (Alpers et al., 2021; Garnett et al., 2021; Neill et al., 2020) , and women significantly increasing the frequency of alcohol use during the pandemic (Pollard et al., 2020) . Three papers found a significant increase in hazardous drinking among women. Barbosa et al. (2020) found that the proportion of people who increased their use to exceed drinking limits during COVID-19 was larger for women than men; although there was no difference in the number of drinks per day. Boschuetz et al. (2020) also found that women were more likely to develop high-risk behavior during social distancing (63% vs. 25%, p = 0.03). A large increase in the number of women who increased their AUDIT score did so based only on the question regarding frequency of alcohol consumption per week. Niedzwiedz et al. (2021) Rodriguez et al. (2020) , found the impact of distress on the maximum number and typical number of drinks consumed per occasion was moderated by gender, and that psychological distress was associated with greater drinking in women but not in men.

Looking at the interaction of age and gender, Callinan et al. (2021) found no overall significant difference between men and women's change in alcohol use, but the interaction between age and gender showed young men decreased their use more than other groups. They also found that women aged 30-50 increased both frequency and J o u r n a l P r e -p r o o f quantity of consumption and were the only group to not decrease their harmful drinking .

Age: Twenty-eight papers looked at the impact of age on SU and reported different trends, with four papers reporting no significant changes based on age (Barbosa et al., 2020; Boschuetz et al., 2020; Sidor and Rzymski, 2020; Szajnoga et al., 2020) . Seven papers reported a decrease in SU related to age. Two papers reported that increasing age was associated with a decrease in drinking (Constant et al., 2020; Vanderbruggen et al., 2020) , and five papers found decreased drinking among people Callinan et al. (2020) found that all ages decreased their risky drinking, except for men aged 26 to 35 years old, with the largest decrease of risky behavior seen among the youngest people.

Race: Ten studies reported the impact of race on SU. Three found a nonsignificant impact on SU (Boschuetz et al., 2020; Niedzwiedz et al., 2021; Wardell et al., 2020) . Four studies found that racialized groups were more likely to increase their SU or Mental Health: In 13 studies, mental health was associated with increased SU (Bartoszek et al., 2020; Chodkiewicz et al., 2020; Czeisler et al., 2020; Dozois, 2021; Garnett et al., 2021; Jacob et al., 2021; Neill et al., 2020; Rolland et al., 2020; Sallie et al., 2020; Stanton et al., 2020; Tran et al., 2020; Wardell et al., 2020; Weerakoon et al., 2020) . Higher rates of depression (Bartoszek et al., 2020; Czeisler et al., 2020; Dozois, 2021; Neill et al., 2020; Stanton et al., 2020; Tran et al., 2020; Wardell et al., 2020; Weerakoon et al., 2020) , anxiety (Czeisler et al., 2020; Dozois, 2021; Garnett et al., 2021; Sallie et al., 2020; Stanton et al., 2020; Tran et al., 2020; Weerakoon et al., 2020) , overall poor mental health (Jacob et al., 2021), and current treatment for a mental disorder (Rolland et al., 2020) , were all found to be associated with increased SU during COVID-19. Wardell et al. (2020) reported that depression was positively associated with greater coping motives for drinking. While Chodkiewicz et al. (2020) found no significant difference in rates of treatment for mental health conditions between low-risk and J o u r n a l P r e -p r o o f hazardous drinkers, people who drank more reported coping less well with everyday functioning, and those who were currently having suicidal thoughts (10.7%) were more likely to report increased alcohol use. (2020) found that alcohol use most frequently increased among people who reported having an "alcohol addiction." Weerakoon et al. (2020) found that those who reported ever binge drinking reported a greater increase in drinking, Neill et al. (2020) found that heavy drinking prior to the pandemic was associated with increased COVID-19-related drinking, and Boschuetz et al. (2020) found that people drinking 5-6 drinks daily before the pandemic had more significantly increased use during the pandemic.

Caregiving responsibilities were associated with increased pandemic SU in nine studies (Boschuetz et al., 2020; Callinan et al., 2021; Czeisler et al., 2020; Gadermann et al., 2021; Grossman et al., 2020; Rodriguez et al., 2020; Rossinot et al., 2020; Sallie et al., 2020; Vanderbruggen et al., 2020) , and less reduction in SU in another (Zajacova et al., 2020) . Having children at home was related to increased overall consumption (Callinan et al., 2021; Gadermann et al., 2021; Grossman et al., 2020; Rodriguez et al., 2020; Rossinot et al., 2020; Sallie et al., 2020; Vanderbruggen et al., 2020) , higher AUDIT scores (Boschuetz et al., 2020; Sallie et al., 2020) , number of drinks per occasion (Rodriguez et al., 2020) , number of drinks consumed weekly (Sallie et al., 2020), and J o u r n a l P r e -p r o o f greater number of heavy drinking episodes (Rodriguez et al., 2020) . The odds of consuming more alcohol during lockdown also increased as the number of children at home increased (OR = 1.22 for every one child [CI: 1.15, 1.30], p <0.001) (Vanderbruggen et al., 2020) . Having a child under 18 at home was also associated with more coping-related motives for drinking alcohol (Wardell et al., 2020) . Among Americans who did not report SU in May 2020, unpaid caregivers for adults had 3.33 times the odds of reporting having started to use substances than others to cope with stress in June 2020 (CI: 1.75, 6.31, p<.001; Czeisler et al., 2020) . Contrary to these results Chodkiewicz et al. (2020) found that people with children were more likely to reduce than increase their alcohol use and Weerakoon et al. (2020) found that households with children at home had 0.74 lower odds of binge drinking compared with households with no children (95% CI: 0.58, 0.94)

The impact of employment also demonstrated mixed results. No significant differences were found in six studies by employment status (Chodkiewicz et al., 2020; Rolland et al., 2020; Sidor and Rzymski, 2020; Villanueva et al., 2021; Weerakoon et al., 2020; Zajacova et al., 2020) . Income loss, job loss, economic worries, and unemployment were found to be associated with increased rates of SU in five studies (Alpers et al., 2021; Killgore et al., 2021; Neill et al., 2020; Vanderbruggen et al., 2020; Wardell et al., 2020) . Contrary to these results, two studies found unemployment was a protective factor for increased alcohol use (Rossinot et al., 2020; Szajnoga et al., 2020); Czeisler et al. (2020) reported that more employed people compared with unemployed people started or increased their use to cope with stress; and Grossman et al. (2020) reported that people stressed by spending more time at work increased their alcohol use.

Two studies found that working or studying from home was associated with increased alcohol use (Alpers et al., 2021; Szajnoga et al., 2020) , while another did not find a significant effect (Boschuetz et al., 2020) .

infection (Garnett et al., 2021) or worrying about a relative's health was associated with reduced alcohol use (Rossinot et al., 2020) . However, the perceived threat of COVID-19 infection was positively associated with multiple measures of drinking (e.g., frequency, amount, severity; Rodriguez et al., 2020) . Two studies found non-significant impacts of quarantine on alcohol use (Chodkiewicz et al., 2020; Sallie et al., 2020) , while one found it was associated with increased alcohol use (Alpers et al., 2021) and one found those who were self-isolating alone had lower alcohol use than those not in isolation (Newby et al., 2020) . The severity of and amount of time spent under public health measures, such as lockdown or social isolation appears to affect alcohol use. Huckle et al. (2020) found that drinkers consumed 13% more in the more severe Level 3 lockdown restrictions in the UK compared with Level 2, and Killgore et al. (2021) found that AUDIT scores increased month by month as lockdown continued with the lowest AUDIT scores observed during the first month. Conversely, López-Bueno et al. (2020) found that each week of lockdown in Spain was associated with a reduction in alcohol use.

Coping: Ten studies found that psychological stress levels were associated with increased alcohol use (Alpers et al., 2021; Callinan et al., 2021; Grossman et al., 2020; Kriaucioniene et al., 2020; Neill et al., 2020; Oksanen et al., 2021; Rodriguez et al., 2020; Rossinot et al., 2020; Stanton et al., 2020; Taylor et al., 2021) , while two studies found no association (Chodkiewicz et al., 2020; Đogaš et al., 2020) . Garnett et al. (2021) J o u r n a l P r e -p r o o f found that significant stress about becoming sick with COVID-19 was independently associated with both drinking less (adjusted odds ratio (aOR) Types of coping strategies were related to increased alcohol use in three papers. Chodkiewicz et al. (2020) reported that people in the hazardous drinking group more often used denial coping strategies, and Wardell et al. (2020) found that coping motives were associated with increased alcohol consumption in the past 30 days, controlling for pre-COVID-19 alcohol use. Taylor et al. (2021) found motives related to coping were associated with higher AUDIT alcohol abuse and drug abuse scores; 26% of their sample reported they tried to cope with self-isolation by consuming more drugs or alcohol.

The goal of our review was to identify how COVID-19 impacted the distribution of SU-related risks and behaviors. We identified 53 studies published by March 2021 that described the early impacts of COVID-19 on SU at a population level. Our review found evidence that there has been increased frequency, quantity and severity of SU, particularly alcohol use, among certain segments of the population in certain countries.

However, as also corroborated by studies which appeared after our search deadline, in many countries there were more people decreasing than increasing their use (see the large population-based survey studies in 21 European countries: Kilian et al. (2021) and Manthey et al. (2021) ).

The impact of COVID-19 appears to be minimal, or at least variable for the people represented by Tier 1, 2 and 3 in our NBP model. The changes within these groups seem to be dependent on the sub-population and their pre-pandemic drinking patterns (e.g., mostly at home etc.). Multiple studies found that both the proportion not drinking alcohol and the proportion of people drinking heavily increased, while the proportion of people who drank occasionally decreased, which may indicate a polarization based on different mechanisms: reduced availability of alcohol and drinking occasions, together with fewer resources available due to economic constraints may have led to a reduction for many, while increased stress including mental stress may have led to an increase, in particular among people who already consumed at heavy levels (e.g., Rehm et al., 2020) . This seems to be similar to the mechanisms observed in other crises, such as the last great recession (Bor et al., 2013) . Consistent with such mechanisms, Callinan et al. (2021) reported the greatest reduction in alcohol use was among people who drank most of their alcohol outside their home prior to the pandemic.

The clearest negative SU outcomes in the context of COVID-19 are among people categorized as needing Tier 4 and 5 services who have history of problems related to their substance use and concurrent disorders. Across multiple measurements, the studies we reviewed indicated that people who already drank in risky ways before the pandemic were more likely to increase their substance use during the pandemic (see additionally Barrio et al., 2021 published after our search closure). This may have significant implications for service planning. Additionally, as drinking to cope with stressful and negative emotions is associated with risk for developing alcohol-related problems (McPhee et al., 2020; Rodriguez et al., 2020) , the association between coping motives J o u r n a l P r e -p r o o f and increased pandemic alcohol use may also result in an upward shift in drinking patterns among pre-pandemic drinkers at a moderate level of risk. Given the evidence suggesting an increase in SU among those who are more vulnerable to harms and association with psychosocial distress, anxiety and depression, and overall poor mental health, these factors may complicate the process of treatment and support. During the pandemic, although health care visits for all reasons have decreased significantly, visits specifically for mental health and substance use concerns have increased (Yang et al., 2020) . Significant collaboration and coordination among SU and other health and social service sectors, including mental health services are clearly indicated (Rush, 2014) .

Increased substance use during COVID-19 among those currently having suicidal thoughts is an important finding to investigate in future studies (Chodkiewicz et al., 2020) . Having an alcohol use disorder increases the risk of committing suicide, and acute alcohol use is associated with greater risks of suicide attempts (Flensborg-Madsen et al., 2008; Borges et al., 2017) . Although preliminary evidence does not show a significnat increase in suicide rates during the pandemic (Pirkis et al., 2021) , since the use of substances can act as a trigger of suicidal behavior, health systems need to increase screening and offer appropriate prevention and treatment to those having suicidal thoughts (Vijayakumar et al., 2011) .

In order to meet the changing needs of a population, it is important to monitor the impact of disease outbreaks such as COVID-19 and the resulting public health measures, including considerations of gender, race, and socioeconomic status (Rehm et al., 2020; Vigo et al., 2020) . Maintaining a focus on diversity and equity is an essential and too often neglected principle of treatment system planning and evaluation (Rush and J o u r n a l P r e -p r o o f Urbanoski, 2019). In the papers we reviewed, increased SU was associated with caregiving responsibilities. Although not all studies showed a significant impact of gender on increased SU, higher stress is linked to unpaid caregiving for children and sick family members, which is often gendered (Rodriguez et al., 2020) . The stress associated with caregiving and the increased risk of domestic violence during lockdown may lead to increased SU among women and the need for trauma and gender-informed services (Australian Bureau of Statistics, 2020; Finlay and Gilmore, 2020). SU, particularly alcohol use, is also a well-known risk factor for violence perpetration, and domestic violence may be both exacerbated and hidden by the social isolation mandated by public health measures (van Gelder et al., 2020) .

The evidence in our review only includes papers published prior to April 2021.

The vast majority of these studies were conducted within the first month or two of the pandemic and the associated public health measures at that time. It is unclear if these changes in SU will be maintained. There is some evidence that the initial increase at the start of the pandemic has lessened over time (López-Bueno et al., 2020; Wright et al., 2020) , while other studies associate increased substance with longer and more severe public health restrictions (Huckle et al., 2020; Killgore et al., 2021) . In addition to being a clear limitation in answering our research questions, the sheer lack of empirical work related to non-alcohol SU also signals a need for more research and population surveillance on which to base prevention and treatment policies and programming. The 

Our review provides preliminary data on changes in substance use during the COVID-19 pandemic. This indicates that certain segments of the population increased their alcohol use, and may be at greater risk of harm and in need of additional services. In particular, people who already drank in risky ways before the pandemic were more likely to increase their substance use. Caregiving responsibilities, stress, depression, anxiety, and current treatment for a mental disorder were also found to be at associated with increased substance use. There continues to be a need for population-level information on SU and SUD by gender and age to inform evidence-based, rapid responses to the COVID-19 pandemic from a treatment system perspective. Every-Palmer, S., Jenkins, M., Gendall, P., Hoek, J., Beaglehole, B., Bell, C., Williman, J., Rapsey, C., Stanley, J., 2020. Psychological distress, anxiety, family violence, suicidality, and wellbeing in New Zealand during the COVID-19 lockdown: A crosssectional study. PLoS one 15(11), e0241658.

Ferrante, G., Camussi, E., Piccinelli, C., Senore, C., Armaroli, P., Ortale, A., Garena, F., Giordano, L., 2020. Did social isolation during the SARS-CoV-2 epidemic have an impact on the lifestyles of citizens? Epidemiol Prev, 353-362. Exposure to or living in countries affected by the Coronavirus (COVID-19) or the impact of associated public health measures including lockdown, social distancing and stay-athome orders.

Before the pandemic; to another population that did not experience the pandemic event; groups within a population; or no comparison.

Changes substance use service need as identified by changes in prevalence, incidence, or severity of substance use and substance use disorders.

J o u r n a l P r e -p r o o f 

Shelter-at-home orders related to COVID-19 pandemic Self-reported change and motivation for alcohol use 60.1% increased alcohol use compared to pre-COVID.

12.8% reported that their alcohol use had decreased from pre-post COVID-19.

27.0% did not change their alcohol use.

J o u r n a l P r e -p r o o f

Abstinence Among Alcohol Use Disorder Patients During the COVID-19 Pandemic: Insights From Spain

Mental well-being (Depression, loneliness, insomnia, daily life fatigue) during COVID-19 related homeconfinement-A study from Poland

A meta-analysis of acute alcohol use and the risk of suicide attempt

Nutrition behaviors in Polish adults before and during covid-19 lockdown

Alcohol Use During the Great Recession of

Changes in Alcohol Use Patterns in the United States During COVID-19 Pandemic

Purchasing, consumption, demographic and socioeconomic variables associated with shifts in alcohol consumption during the COVID-19 pandemic

Shifts in alcohol consumption during the COVID-19 pandemic: early indications from Australia

COVID-19 and Increased Alcohol Consumption: NANOS Poll Summary Report

Alcohol Consumption Reported during the COVID-19 Pandemic: The Initial Stage

Trauma trends during COVID-19 alcohol prohibition at a South African regional hospital

Socio-cognitive factors associated with lifestyle changes in response to the COVID-19 epidemic in the general population: results from a cross-sectional study in France

Did the General Population in Germany Drink More Alcohol during the COVID-19 Pandemic Lockdown?

Associations between changes in health behaviours and body weight during the COVID-19 quarantine in Lithuania: the Lithuanian COVIDiet Study

SARS-related perceptions in Hong Kong

COVID-19 confinement and health risk behaviors in Spain

Alcohol controls in the aftermath of the COVID-19 pandemic in India

Use of alcohol, tobacco, cannabis, and other substances during the first wave of the SARS-CoV-2 pandemic in Europe: a survey on 36,000 European substance users

Racial and ethnic disparities in the prevalence of stress and worry, mental health conditions, and increased substance use among adults during the COVID-19 pandemic-United States

Depression, Environmental Reward, Coping Motives and Alcohol Consumption During the COVID-19 Pandemic

Alcohol use in Australia during the early days of the COVID-19 pandemic: Initial results from the COLLATE project

Acute mental health responses during the COVID-19 pandemic in Australia

Mental health and health behaviours before and during the initial phase of the COVID-19 lockdown: longitudinal analyses of the UK Household Longitudinal Study

Psychological Stressors Predicting Increased Drinking During the COVID-19 Crisis: A Longitudinal National Survey Study of Workers in Finland

The COVID-19 Pandemic and its Impact on Substance Use: Implications for Prevention and Treatment

Suicide trends in the early months of the COVID-19 pandemic: an interrupted time-series analysis of preliminary data from 21 countries

Changes in Adult Alcohol Use and Consequences During the COVID-19 Pandemic in the US

COVID-19 and mental health: A review of the existing literature

Alcohol use in times of the COVID 19: Implications for monitoring and policy

Drinking to cope with the pandemic: The unique associations of COVID-19-related perceived threat and psychological distress to drinking behaviors in American men and women

Global Changes and Factors of Increase in Caloric/Salty Food Intake, Screen Use, and Substance Use During the Early COVID-19 Containment Phase in the General Population in France: Survey Study

Behavioral Changes During COVID-19 Confinement in France: A Web-Based Study

Tiered Frameworks for Planning Substance Use Service Delivery Systems: Origins and Key Principles

Evaluating the Complex: Alternative Models and Measures for Evaluating Collaboration among Substance use Services with Mental Health

Increase in prevalence of current mental disorders in the context of COVID-19: analysis of repeated nationwide cross-sectional surveys

Global Drug Survey

Second, third, fourth COVID-19 waves and the 'pancession': We need studies that account for the complexities of how the pandemic is affecting alcohol consumption in Australia. Drug and alcohol review

Disparities in outpatient visits for mental health and/or substance use disorders during the COVID surge and partial reopening in Massachusetts

Changes in health behaviours during early COVID-19 and socio-demographic disparities: a cross-sectional analysis

Search Strategy Problem or Patient or Population 1. addiction.mp. or substance abuse.mp. or exp Substance-Related Disorders/ or drug dependance

Opioid/ or exp Heroin Dependence/ or exp Heroin/ or heroin.mp. or methadone.mp. or exp Methadone/ or fentanyl.mp. or exp Fentanyl/ 4. marijuana.mp. or cannabis.mp. or exp Cannabis/ 5. cocaine.mp. or exp Cocaine-Related Disorders/ or exp Cocaine/ or exp Cocaine Smoking

Exposure 7. pandemic.mp. or exp Pandemics/ or exp Coronavirus/ or exp Coronavirus Infections/ or Coronavirus.mp. or exp Middle East Respiratory Syndrome Coronavirus/ or SARS virus.mp. or exp SARS Virus/ or exp Severe Acute Respiratory Syndrome/ or H1N1.mp. or Influenza A Virus, H1N1 Subtype or covid.mp Outcome of Interest 8. exp Prevalence/ or Prevalence.mp. or incidence.mp. or exp Incidence/ or severity

/ 10. service planning.mp. or exp Health Planning/ 11. exp Substance Abuse Treatment Centers/ or addiction services.mp. or treatment.mp. or therapy.mp. or counselling.mp. or exp Counseling/ 12. opioid agonist treatment.mp. or exp Opiate Substitution Treatment/ 13. exp Substance Withdrawal Syndrome/ or withdrawal management

This review was conducted as part of a project that received funding from Health Canada. The views represented here do not necessarily represent those of Health Canada.

The impact of COVID-19 on substance use